What is the formula for the search toolbar?
Our employees have been increasingly frustrated with the results that come up with the search toolbar and I am looking for a way to help educate them on how the quick results decides which content to put on top and then also how the main results page is organized.
Also, how does relevance play into this and how is this determined? And is it teachable over time? Or how can we better impact the results we get from search?
Can you provide a little more info about how your users are finding that the search feature is not meeting their needs? Also, have you checked out the improvements we've made to search in SBS 3.0? Try out searching for some topics on this site for a sample.
As for your other questions, they'll require me to look under the covers for a little more info about how the code works. I'll take a look and get back to you.
The quick search is only intended to give you the top five results in Content, People, and Places. The feature is designed so that users can get a quick sampling of the available search results, then click on the View All Results link for a more comprehensive list. We found this to be the best possible design for balancing speed and effectiveness in our search functionality.
Hi Karl,
I understand that it's designed to be only the top results (didn't know
that it was top 5, so that helps), but if quick search is designed for
the top five results, why if when we test a unique tag that is only used
on one document, does that result not show up in the quick results?
Nova Newcomer
COMMUNITY MANAGER
DIR 503-575-4109
I tested this out on both 2.5.7 and 3.0.4, and in both versions it worked as expected.
Does the All Results search return the expected result? It is possible that your search index is incomplete, and needs to be updated. You can update/rebuild your search index on the System > Settings > Search page.
Hi,
It appears that this may be a limitation with search in 2.5. I have confirmed that this works in 3.0, and in fact you can confirm this by using the search box on this page.
However, I have one more thing to try:
Regards,
Karl
Our search formula is based on the standard for Apache Lucene: http://lucene.apache.org/java/2_4_0/scoring.html. In addition, we give extra weight to search results based on popularity (how often is content viewed, edited or replied to), age, subject, and tagging. For exact details on how this works, you're more than welcome to dig into the code base
.
No problem, I am glad to help. I think that based on your feedback here, you will be very happy with the level of quality in search in SBS 3.0. Refining search was one of the "under-the-radar" improvements in the new release, and from all the responses we've seen so far, it seems to have succeeded.
Regards,
Karl
Jive combines collaboration software, community software & social networking software into the leading SBS solution.
© Copyright 2000–2009 Jive Software. All rights reserved.
915 SW Stark St., Suite 400, Portland, OR 97205